On Wed, 2023-06-14 at 11:28 +0200, Alexander Kanavin wrote: > This was done in a selftest, but that is too late and creates > friction in integration as errors are not seen until autobuilder fails. > > Bonus fix: SUMMARY check wasn't even working, as in the absence > of one set in the recipe there is a default value set from bitbake.conf. > > I left DESCRIPTION check out for now, as many recipes don't actually > have it, and it's set from SUMMARY (plus a dot) if absent. > > Signed-off-by: Alexander Kanavin <[email protected]> > --- > meta/classes-global/insane.bbclass | 26 ++++++++++++++++ > meta/lib/oeqa/selftest/cases/distrodata.py | 36 ---------------------- > 2 files changed, 26 insertions(+), 36 deletions(-)
I think this change is in the right direction but there are a few things I think we need to tweak here. a) I don't like the " if not '/meta/recipes-'" check at all. We now have the groundwork for layer overrides in bitbake so I'd like to complete that, then we can make some of these WARN_QA entries as layer specific. This has the added bounus that other layers and more easily adopt the tests and raise the bar on quality. b) I really really don't want a do_fetch postfunc doing something that isn't fetching. I know why you've done this, so it appears early in the build and so on but it doesn't belong there. There are other sanity tests at parse time in anonymous python which also bother me since the increase parsing time and also shouldn't be there. Those often aren't gated on ERROR_QA/WARN_QA options either so can't be configured. Given all this, do we want to consider a new task which covers these kinds of tests? A new task does have certain overhead but I'm starting to think we might need to do that. Of course there is then a dillema about which task it would need to run before. c) There is an issue that warnings are not restored from sstate. If you rerun this build after seeing the warnings, the warnings will not show again. This is going to confuse users and cause issues to get missed on the autobuilder. There is an open bug for this one, it isn't an easy fix unfortunately. Cheers, Richard
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#182803): https://lists.openembedded.org/g/openembedded-core/message/182803 Mute This Topic: https://lists.openembedded.org/mt/99524137/21656 Group Owner: [email protected] Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
